The innate and early immune response to pathogen challenge in the female genital tract and the pivotal role of epithelial cells.

The female reproductive tract is immunologically unique in its requirement for tolerance to allogeneic sperm and, in the upper tract, to the conceptus. However, it must also be appropriately protected from, and respond to, a diverse array of sexually transmitted pathogens. Some of these infections can be lethal (e.g. Human Immunodeficiency Virus (HIV), Human Papilloma Virus (HPV)), and others (e.g. Chlamydia trachomatis and Neisseria gonorrhoeae) can have potentially devastating reproductive sequelae. Interactions between a host and a pathogen are complex, diverse and regulated, and are a function of the individual pathogen, and host immunity. Although there is undoubtedly commonality in the mucosal immune response, there is also evidence of a degree of site-specificity in immune mechanisms, dependent upon the function and anatomical location of an organ. In this article, we review the evidence on the pivotal role of epithelial cells in the innate and early immune response to pathogen challenge in female genital tract tissues, and examine the evidence that the 'sterile' upper and the 'non-sterile' lower female genital tract may maintain a different immunological surveillance milieu, and may also respond differentially to pathogen challenge. We also review the unique characteristics, and subsequent ramifications of the acute cervical immune response to C. trachomatis, and discuss how natural antimicrobial mediators of immunity may be utilized to decrease the spread of sexually transmitted infections.

Immunization of mice with peptomers covalently coupled to aluminum oxide nanoparticles.

Subunit vaccines generally require adjuvants to elicit immune responses, but adjuvants may alter the conformation of critical epitopes and reduce vaccine efficacy. We therefore tested an immunization strategy in which antigen is covalently coupled to aluminum oxide nanoparticles using a method that favors preservation of the native conformation. The test antigen consisted of "peptomers" (head-to-tail-linked peptide homopolymers) derived from the 4th conserved region (C4) of HIV-1 gp120 which is believed to be in an alpha-helical conformation prior to binding to CD4. Immune responses in mice to peptomer-nanoparticle conjugates were compared to responses elicited by free C4 peptide and C4 peptomers, with and without the hydrophilic adjuvant muramyl dipeptide (MDP). Highest peptomer-specific serum antibody responses were induced by peptomer-particles without MDP. Serum antibodies induced by peptomer-particles also showed highest reactivity towards recombinant, glycosylated gp120 and HIV-1 infected T cells. The results suggest that this novel vaccine approach could be useful for induction of immune responses against conformation-sensitive viral antigens without the need for additional adjuvants.

Detection of HIV-1-specific CTLs in the semen of HIV-infected individuals.

CTLs play an important role in controlling cell-associated HIV. Since the majority of HIV infections are acquired through sexual transmission, we investigated whether antiviral CTLs were present in the male urogenital tract using semen as a source of T cells. We were able to establish anti-HIV cytolytic lines in five of five HIV-infected men with CD4 counts of >500/microl, although cloning efficiencies were lower than with peripheral blood-derived T cells. CTLs generated from the semen of three men were analyzed in detail and showed a broadly active response, recognizing gag, env, and pol proteins. Detailed analysis of two gag-specific clones from one of the individuals demonstrated HLA class I restriction and recognition of the same p24 epitope (EQASQEVKNWMT). In summary, our results demonstrate the presence of a broad CTL response to HIV in the urogenital tract and provide a rationale for further studies of local enhancement of genital mucosal responses by anti-HIV immunization.

Human beta-defensin-1: an antimicrobial peptide of urogenital tissues.

Antimicrobial peptides are widely distributed mediators of innate host defense in animals and plants. A 36 amino acid antimicrobial peptide belonging to the defensin family, and named human beta-defensin-1 (HBD-1), was purified recently from hemodialysate fluid, but its tissue sources were not identified. By Northern blotting, we found the highest concentrations of HBD-1 mRNA in the kidney and the female reproductive tract. In situ hybridization localized the HBD-1 mRNA in the epithelial layers of the loops of Henle, distal tubules, and the collecting ducts of the kidney and the epithelial layers of the vagina, ectocervix, endocervix, uterus, and fallopian tubes in the female reproductive tract. Using a novel technique designed to detect cationic peptides in urine, we recovered several forms of HBD-1 ranging in length from 36 to 47 amino acid (aa) residues and differing from each other by amino terminal truncation. The total concentration of HBD-1 forms in voided urine was estimated at 10-100 microg/liter, with individual variations in the total amount of HBD-1 peptides and the relative proportion of HBD-1 forms. Multiple forms of HBD-1 (size 36-47 aa) were also found in the blood plasma, bound to carrier macromolecules that released the peptide under acid conditions, and in vaginal mucosal secretions (39, 40, and 44 aa). By immunostaining, HBD-1 was located in the kidney within the lumen of the loops of Henle, but no intracellular storage sites were identified in renal or female reproductive tissues. Recombinant HBD-1 forms (36, 39, and 42 aa) and natural HBD-1 forms were antimicrobial to laboratory and clinical strains of Escherichia coli at micromolar concentrations. HBD-1 activity was not changed appreciably by low pH, but was inhibited by high salt conditions. Some of the HBD-1 peptides retained their activity against E. coli in unconcentrated (low conductance) urine, and the 36 aa form was microbicidal even in normal (high conductance) urine. Production of HBD-1 in the urogenital tract could contribute to local antimicrobial defense.

Gene expression, immunolocalization, and secretion of human defensin-5 in human female reproductive tract.

This study describes the novel localization of the antimicrobial peptide human intestinal defensin-5 (HD-5) in female genital tract epithelia. Using a 3' rapid amplification of cDNA ends (RACE) protocol, HD-5 was cloned from a vaginal epithelial cell RNA preparation, and its identity was confirmed by sequencing. Tissue samples from multiple donors were subsequently screened for HD-5 expression by reverse transcription polymerase chain reaction. HD-5 message was invariantly expressed by normal vagina and ectocervix and inflamed fallopian tube, but variably expressed by normal endocervix, endometrium, and fallopian tube (60, 64, and 29% of specimens, respectively). Expression in endometrium was the highest during the early secretory phase of the menstrual cycle. Using immunohistochemistry and confocal microscopy, HD-5 peptide was localized in the upper half of the stratified squamous epithelium of the vagina and ectocervix, with the intensity of cellular staining increasing toward the lumen. In positive endocervix, endometrium, and fallopian tube specimens, HD-5 was located in apically oriented granules and on the apical surface of a proportion of columnar epithelial cells. Using Western blot analysis, secreted HD-5 was detected in cervicovaginal lavages, with the highest concentrations found during the secretory phase of the menstrual cycle. We hypothesize that HD-5 is an intrinsic component of the female urogenital innate immune defense system and that its expression may be modulated by hormonal and proinflammatory factors.

The case against an association between HIV-1 and sperm: molecular evidence.

In this article we present data from our laboratory, and review the literature available, on the potential association between HIV-1 and sperm. We focus on the use of PCR technology to answer this very important question, and emphasise the importance of using highly purified sperm preparations. We conclude that the likelihood of HIV infection/association with viable mature sperm is low.

T lymphocytes and macrophages, but not motile spermatozoa, are a significant source of human immunodeficiency virus in semen.

The cellular fraction of semen contains spermatozoa, immature germ cells, leukocytes, and epithelial cells. Recent evidence implicates seminal cells as a major source of sexually transmitted human immunodeficiency virus (HIV) in semen, but the identity and infectious potential of infected cells remains poorly understood. HIV provirus was found in 75% of viable semen cell samples by polymerase chain reaction and in 88% of paired blood cell samples from HIV-seropositive men. When semen cell subpopulations were isolated by an immunomagnetic bead technique, T cells were found to be most commonly HIV-infected (75% of samples), followed by macrophages (38% of samples). Viral DNA was never detected in motile spermatozoa or immature germ cell populations. Semen leukocytes proliferated in response to mitogenic and antigenic challenge and produced p24 following stimulation with irradiated allogeneic cells. These data provide evidence that both T cells and macrophages, but not germ cells, are cellular vectors of HIV transmission in semen.

Optimized assay for antisperm cell-mediated immunity.

Previous studies on antisperm cell-mediated immunity (CMI) have been confounded by the presence of immunogenic leukocytes in sperm antigen preparations. In this study we isolated pure populations of viable spermatozoa on discontinuous Percoll gradients, and utilized sonicated and cavitated extracts, as well as live motile spermatozoa, to measure cellular immunity to spermatozoa in vasectomized men, men with proven fertility, infertile women, fertile women and umbilical cord blood. Using a thymidine incorporation assay to assess lymphocyte proliferation, nine out of 13 (69%) vasectomized men and five out of 10 (50%) fertile men responded to sperm extracts. Lymphocyte proliferation to sperm extracts was also observed in both infertile and fertile women (27 and 50% respectively). In addition, viable sperm preparations promoted lymphocyte responses in five out of eight (63%) fertile women, seven out of 11 (63%) healthy men and four out of 11 (45%) cord blood specimens. Furthermore, four out of 11 (36%) healthy normal men responded to autologous spermatozoa. No relationship between serum antisperm antibodies, as measured with the Immunobead test, and sperm CMI was observed in any group. This study provides evidence that lymphocytes from fertile as well as infertile men and women and sperm-naive newborn infants proliferate when exposed to viable spermatozoa or sperm extracts. Thus the lymphocyte proliferation assay does not appear to be useful in the diagnosis of immunological infertility, but immunological recognition of spermatozoa may be a common feature that could have a role in fertility.

Secretion of proinflammatory cytokines by epithelial cells in response to Chlamydia infection suggests a central role for epithelial cells in chlamydial pathogenesis.

Chlamydia species infect epithelial cells at mucosal surfaces, and are major causes of sexually transmitted diseases. Infection is characterized by inflammation which is exacerbated upon reinfection, ultimately leading to tissue damage and scarring. Although central for the development of disease manifestations, little is known about the mechanisms that initiate and sustain the inflammatory response to Chlamydia. Infection of cervical and colonic epithelial cells with Chlamydia trachomatis and Chlamydia psittaci is shown in the present studies to upregulate mRNA expression and secretion of the proinflammatory cytokines IL-8, GRO alpha, GM-CSF, and IL-6. In contrast to the rapid, but transient, cytokine induction following infection with other invasive bacteria, the epithelial cytokine response to Chlamydia was delayed until 20-24 h after infection, persisted throughout the chlamydial growth cycle (2-4 d), and required bacterial protein synthesis. Moreover, epithelial cell lines and primary endocervical epithelial cells released IL-1alpha after Chlamydia infection, and increased secretion of the proinflammatory cytokines could be inhibited by anti-IL-1alpha. This suggests that IL-1alpha, released following lysis of infected epithelial cells, may amplify the inflammatory response by stimulating additional cytokine production by noninfected neighboring cells. These findings suggest a novel pathophysiologic concept wherein the acute host response to Chlamydia at mucosal surfaces is primarily initiated and sustained by epithelial cells, the first and major targets of chlamydial infection.

Superantigen-mediated proliferation and cytotoxicity of T cells isolated from the inflammatory tissues and peripheral blood of arthritis patients.

Superantigens are thought to play a role in acute infections and in the pathogenesis of autoimmune diseases that are believed to have an infectious etiology. The effect of the superantigens staphylococcal enterotoxin A, staphylococcal enterotoxin B, and streptococcal M type 5 protein on T cells derived from inflammatory tissues and peripheral blood (PB) of arthritis patients was studied in seven rheumatoid arthritis (RA), two psoriatic arthritis, two reactive arthritis, and one ankylosing spondylitis patient. Superantigen-reactive T cells and T cell lines derived from the PB, synovial fluid (SF), and synovial membrane (SM) of all 12 arthritis tissues recognized the superantigens in an MHC-unrestricted manner. Heterogeneities in proliferation and superantigen-directed T cell cytotoxicity were observed in E+ T cells and the T cell lines. Four SF-CD4+ mycobacteria heat-shock protein 65-kDa specific T cell clones generated from an RA patient could recognize and lyse each other when pulsed with staphylococcal enterotoxin A and used as targets. From another RA patient, four SF-CD4+ T cell clones that specifically recognize autoantigens were generated with human IgG fragments or collagen type II fragments. Heterogeneities of such superantigen-mediated specific lysis were also demonstrated. The data presented by us suggest a model in which superantigens do not have to be involved in triggering the initial disease because autoreactive T cells elicited by antigen can, in the presence of superantigen, lyse cells that express MHC class II molecules, including activated T cells.

Characterization of T lymphocytes and antigen-presenting cells in the murine male urethra.

The male lower urogenital tract is exposed to sexually transmitted pathogens and is therefore a strategic site of immune defense. To further define the immunodynamics of this region, we studied the histology, immune cell distribution, and draining lymph nodes of the murine male lower urogenital tract. The external surface of the foreskin was covered by skin composed of keratinized stratified epithelium containing numerous hair follicles and sebaceous glands. Immunologically the penile foreskin was characterized by the presence of few T lymphocytes and macrophages. Numerous Langerhans cells, however, were detected within the epithelium. The penile urethra was composed of stratified columnar epithelium, with a meatus lined by keratinized squamous epithelium preceding the opening proper. The most abundant immune cells of the penile urethra were macrophages. In young adult, virgin males, these were found primarily underlying the urethral epithelium, but in older, mated mice, they were usually intraepithelial in location, and were more abundant. Langerhans cells could not be specifically identified in the urethral mucosa. T lymphocytes were found underlying and occasionally within the epithelium of the urethral mucosa, with CD4+ cells more abundant than CD8+ cells. The majority of lymphocytes observed around the urethra were positive for the integrin beta 7 alpha M290, which is selectively expressed by mucosal lymphocytes, providing indirect evidence that the urethra is part of the common mucosal system. Lymphocytes expressing the gamma delta T cell receptor and IgA-positive plasma cells were not detected. The primary draining nodes for the vas deferens and urethra were the lumbar nodes. Lymphatic drainage from the rectum also involved the lumbar nodes. Information obtained in this study should help to elucidate optimal genital tract vaccination strategies for defense of the male urogenital tract against sexually transmitted pathogens.

Different cytokine production profiles of gamma delta T cell clones: relation to inflammatory arthritis.

This study was performed to investigate whether gamma delta T cells could also be divided into subsets, identified by a cytokine profile, as described for alpha beta T helper (Th) cell subsets. Cytokine production was studied in 22 gamma delta T cell clones obtained from the synovial fluid and peripheral blood of one patient with inflammatory arthritis and compared to that of 26 alpha beta T cell clones of the same and different patients. Interferon-gamma (IFN-gamma) was produced by 18 (82%) and interleukin-4 (IL-4) by 17 (77%) out of 22 gamma delta T cell clones, respectively. In contrast, IL-10 was not produced, except at very low level in one case. The mean levels of IL-4 were lower for clones derived from synovial fluid. When considering the production of IFN-gamma as an indicator of Th1 and that of IL-4 as an indicator of Th2, respectively, the most common pattern was a gamma delta Th1-like pattern, with the combination of high levels of IFN-gamma and low levels of IL-4. This pattern was found in V delta 1+ clones, all from synovial fluid. Additional patterns were also observed: a mixed, probably gamma delta Th0-like pattern with a more balanced production of both IFN-gamma and IL-4; a gamma delta Th1 pattern with the production of IFN-gamma alone; a gamma delta Th2 pattern with the production of IL-4 alone. These three patterns were also seen in blood gamma delta T cells which were all V delta 2, indicating that these patterns were independent of the V delta phenotype. gamma delta T cell clones produced lower levels of IFN-gamma (p = 0.001) and higher levels of IL-4 than alpha beta clones (p < 0.02). These differences in cytokine production between alpha beta and gamma delta subsets and within these subsets may contribute to their respective role in chronic inflammation.

Selective activation of resting human gamma delta T lymphocytes by interleukin-2.

In rheumatoid arthritis and other inflammatory diseases we and others have found that gamma delta T cells express activation antigens, suggesting that they are involved in the pathogenesis of these disorders. In this study we have stimulated peripheral blood mononuclear cells from normal donors with recombinant interleukin-2 (rIL-2) to see whether such a stimulus alone could activate gamma delta T cells. Short-term exposure (24-96 h) to rIL-2 selectively stimulated the gamma delta but not the alpha beta T cells to express activation antigens (CD69, CD25 and HLA-DR). Long-term culture (2 weeks) in rIL-2-containing medium caused a selective increase in the proportion of the gamma delta T cells and a corresponding reduction of the fraction of alpha beta T cells. Limiting dilution analysis revealed that approximately 1/60 of the gamma delta T cells responded to IL-2 in contrast to only 1/250 of the alpha beta T cells. Comparison of the expression of the IL-2 receptor (IL-2R) alpha and beta chains showed that there was a similar expression of the alpha chain on gamma delta and alpha beta T cells whereas the relative density of the beta chain was more than twice as high on gamma delta T cells. Both the IL-2-induced proliferation of gamma delta T cells and the expression of activation antigens on these cells could be inhibited by an anti-IL-2R beta monoclonal antibody (mAb) but not by an anti-IL-2R alpha mAb. Expression of CD69 on gamma delta T cells was dependent neither on the presence of B cells, monocytes, nor alpha beta T cells. Finally, we found that the IL-2-induced expression of CD69 was inhibited by activation of cAMP-dependent protein kinase and by inhibition of the Src-family of the tyrosine protein kinase, but not by inhibition of protein kinase C or by activation of the CD45 associated tyrosine phosphatase. The ability of gamma delta T cells to be activated by IL-2 is a feature which they have in common with natural killer cells. Moreover, it may be possible that the expression of activation antigens on gamma delta T cells in inflammatory diseases is an epiphenomenon secondary to IL-2 produced by activated alpha beta T cells.

Rheumatoid inflammatory T-cell clones express mostly Th1 but also Th2 and mixed (Th0-like) cytokine patterns.

This study was performed in order to characterize whether T cells from rheumatoid synovial inflammation belong to the Th1- or Th2-like functional subsets. Cytokine production was studied in 26 CD4+ alpha beta+ and 2 CD8+ alpha beta T-cell clones from the synovial fluid, the synovial membrane and peripheral blood of 5 patients. Fifteen of the CD4+ clones were raised against various mycobacterial antigens and 11 CD4+ clones and 2 CD8+ clones were raised unspecifically using PHA and/or IL-2. The specificities of these clones are not known. In the mycobacterial antigen-specific group, all CD4+ alpha beta T-cell clones produced IFN-gamma at high levels, while the production of IL-4 was generally absent or low (< 1 ng/ml), consistent with a Th1-like profile. Some of these clones, however, also produced various amounts of IL-10 which has been regarded as a Th2 product but can be produced also in lower amounts by Th1 cells. One HSP-65-specific clone produced levels of IL-4 and IL-10 in the same order as that of IFN-gamma, thus appearing to be Th0-like. Among the 11 unspecific CD4+ clones, 7 showed a Th1-like pattern but with lower levels of IFN-gamma than the antigen-specific clones. However, three clones did not produce any IFN-gamma activity but produced IL-4 and one of them also produced distinct amounts of IL-10, compatible with a Th2-like pattern. In addition, one of the clones also showed an almost equally strong IFN-gamma and IL-4 production, thus most likely representing a Th0-like clone.

Heterogeneity of the TCR repertoire in synovial fluid T lymphocytes responding to BCG in a patient with early rheumatoid arthritis.

T lymphocytes have been implicated in the pathogenesis of rheumatoid arthritis. Interestingly, many of the activated T cells isolated from the synovial fluid of individuals with rheumatoid arthritis react with antigens from Mycobacterium tuberculosis or BCG. This response is seen to a much lesser extent in the peripheral blood of these patients. To investigate the nature of the T-cell response to BCG in RA, we isolated T cells from the synovial fluid of a patient with early-stage rheumatoid arthritis, stimulated them with BCG and cloned by limiting dilution. Staining with monoclonal antibodies specific for different V beta gene families revealed a statistically significant greater proportion of synovial-derived T-cell clones expressing the V beta 8 gene family product compared with peripheral blood clones. While the antigen specificity of some of the clones could not be determined, several of the clones displayed distinct antigen reactivities. Sequencing the TCR beta chain genes of these T cells suggested that although the V beta 8 gene products appeared to be over-represented in these BCG-specific clones, each clone utilized distinct J beta gene segments and used N segment addition to different extents. In addition, no common motifs were identified in the beta chain CDR3s of the clones sequenced. Analysis of bulk cultured BCG-specific SF T cells and unstimulated peripheral blood T cells for V beta 8 gene expression also revealed a large amount of diversity within the CDR3 region. Thus, the T-lymphocyte response to BCG in this patient with early rheumatoid arthritis appears to be quite heterogeneous.

A high proportion of the V delta 1+ synovial fluid gamma delta T cells in juvenile rheumatoid arthritis patients express the very early activation marker CD69, but carry the high molecular weight isoform of the leucocyte common antigen (CD45RA).

We have previously shown that gamma delta T cells in the synovial compartment of patients with juvenile rheumatoid arthritis (JRA) express activation antigens (CD69 and HLA-DR) and that they are predominantly of the V delta 1 subset. In this study we have analysed the expression of activation antigens (CD69 and HLA-DR) and different isoforms of the leucocyte common antigen (CD45RO and CD45RA) on the V delta 1 and the V delta 2 subsets of gamma delta T cells in paired samples of synovial fluid and peripheral blood of nine patients with JRA, and in the peripheral blood of five children with idiopathic scoliosis. In the synovial fluid of children with JRA, there were significantly more V delta 1+CD69+ and V delta 2+CD69+ cells compared with the peripheral blood of the same patients. In contrast, however, in the synovial fluid the V delta 1 and the V delta 2 subsets differed with respect to the expression of the two isoforms of the leucocyte common antigen. The majority of the V delta 1+ cells expressed the high molecular weight isoform (CD45RA+) while most of the V delta 2+ cells carried the low molecular weight variant (CD45RO+) of this molecule.

TCR gamma/delta+ cell subsets in the synovial membranes of patients with rheumatoid arthritis and juvenile rheumatoid arthritis.

Using a peroxidase/anti-peroxidase immunohistochemical staining method, we examined sections of inflammatory synovial membranes from 13 patients with juvenile rheumatoid arthritis (JRA) and 11 with rheumatoid arthritis (RA). The relative numbers of TCR gamma/delta+ cells and the proportions of V delta 1+ and V delta 2+ subsets were recorded in the areas of the membranes most heavily infiltrated by CD3+ cells. In the JRA group, the majority (8/13) of the membranes had TCR gamma/delta+ cells which contributed between 5 and less than 10% of the total number of CD3+ cells. In the RA synovial membranes examined, 5/11 samples had between 5 and 10% TCR gamma/delta+ cells, but in another 5 TCR gamma/delta+ cells contributed to between 10 and 20% of CD3+ cells. No significant difference was noted between the two patient groups. However, the range of values found in the RA membranes appeared to be slightly higher in comparison to previously reported values for RA synovial fluid, peripheral blood and eluted synovial membrane T cells. Analysis of the relative proportions of the V delta 1+ and V delta 2+ subsets revealed a significant dominance of V delta 1+ cells in RA membranes and approximately equal numbers of the two populations in the JRA patients. As the majority of peripheral blood TCR gamma/delta+ cells use the V delta 2 segment this suggests a preferential homing or expansion of the V delta 1+ cells in both RA and JRA synovium. The overall distribution pattern of the TCR gamma/delta+ and V delta 1+ and V delta 2+ cells was also recorded. These cells mostly accumulated in the lymphoid-like tissues and in the perivascular area in the tissues of both RA and JRA patients. Occasionally, augmented numbers of these cells were found in the subsynovial layer or in the loose connective tissue. In the majority of cases, only a few TCR gamma/delta+ cells were located in the synovial layer. The function and the possible pathogenetic importance of these TCR gamma/delta+ cells have not so far been determined.

Peptide recognition, T cell receptor usage and HLA restriction elements of human heat-shock protein (hsp) 60 and mycobacterial 65-kDa hsp-reactive T cell clones from rheumatoid synovial fluid.

A commonly held postulate regarding the etiology of rheumatoid arthritis (RA) is that of antigenic mimicry. Recent interest has focused on the mycobacterial 65-kDa heat-shock protein (hsp) as a putative causal agent. The 65-kDa hsp has over 40% sequence homology with the human hsp 60, and elevated synovial T cell responses to both antigens have been demonstrated in RA and juvenile rheumatoid arthritis patients. Such T cells should, therefore, be specific for shared epitopes on the two antigens. To investigate this, we screened synovial fluid mononuclear cells from two early RA patients with peptides of the 65-kDa hsp which have the greatest homology with the human hsp 60. We also raised a panel of T cell clones from one of the patients with the 65-kDa hsp. The synovial T cell population from both patients and one of the T cell clones recognized a peptide representing the amino-acid sequence 241-255. This clone also responded to the peptide of the equivalent human sequence, and was restricted by HLA-DQ. A second T cell clone recognized an adjacent epitope (amino acid sequence 251-265) which is also highly homologous with the human sequence, but this clone was restricted by HLA-DR. The clones utilized different V beta gene segments but the same D beta and J beta gene elements, and both exhibited specific cytotoxicity against autologous antigen-pulsed macrophages. Our findings, therefore, do not disagree with the postulate that autoimmune disease could possibly be triggered by bacterial epitopes with homology to self protein. However, it is also noted that there are alternative interpretations of this data.